This K02 renewal proposal describes three major aims: To develop new statistical methods to better answer substantive questions in alcohol research related to the development and prevention of alcohol problems, to collaborate with alcohol researchers on advanced statistical analysis of their data, and to disseminate information on the new methods so that others in the alcohol fields can benefit from them. Five areas of statistics integral to the study of the development and prevention of alcohol problems will be studied: mixture data analysis, categorical and other non-normal data analysis, multilevel data analysis, missing data analysis, and longitudinal analysis. Within the area of longitudinal analysis three topics will be studied: Extensions of survival analysis, randomized trials, and power and design issues. Collaboration with researchers on the analysis of their alcohol data will involve advanced analyses of several longitudinal data sets concerned with alcohol problems. Dissemination of information on the new methods developed within the project will take place through the publication of substantively- and methodologically-oriented articles, presentations at alcohol and statistical meetings, training sessions, visits at alcohol research centers, and hosting visitors at my UCLA methods group. Two major types of career development and training activities are planned within the K02: Training on substantive research issues related to the development and prevention of alcohol problems and advanced statistical training. The central part of the substantive training will occur through collaboration with a few key experts in alcohol and related fields.